Almost half of the world's population uses solid fuels for cooking. In low-resource countries, pregnant women and their neonates are exposed to dangerously high levels of household air pollution from solid fuel smoke. In 2010, household air pollution caused an estimated 4 million deaths globally, including 100,000 from neonatal pneumonia. This is likely an underestimate of global burden because it does not include prenatal exposures to household air pollution and association with preterm birth. Despite best efforts around the world to develop clean-burning solid fuel stoves, very few stoves have consistently reduced household air pollution to levels that are needed to protect health. Only electric and gas stoves have that potential, but electric stoves are not a viable option in regions with limited access to reliable electrical power. Liquid petroleum gas (LPG) stoves can dramatically reduce household air pollution when fully adopted, however, the level of adoption and sustained use of gas stoves under real-life conditions is unknown. Furthermore, other exposures to outdoor and indoor cooking fires and tobacco smoke may occur, and tailored interventions to avoid these exposures have not been tested. We propose to assess the feasibility and acceptability of an introduced gas stove complemented by two alternative approaches to delivering a behavioral intervention using a Social Cognitive Theory-based framework. In the first approach, 30 pregnant women will receive a gas stove and will be taught by peer educators in group classes how to safely use gas stoves and how to reduce exposure to other sources of air pollution. In the second approach, 30 additional pregnant women will receive the gas stove and group classes, but in this phase, peer educators will visit each home after the classes to help the women and their families identify top priority strategies to reduce ar pollution. Using a checklist, peer educators will observe whether household members adhere to tailored strategies and provide ongoing support to families. We will use stove use monitoring devices to assess gas stove uptake and sustained use, measure maternal urinary metabolites of polycyclic aromatic hydrocarbons before and after gas stove introduction and measure perinatal and neonatal exposures to carbon monoxide and particulate matter during the successive behavioral change interventions. Health assessments and household air pollution measurements will be conducted on pregnant women and their neonates. Neonates will be assessed for preterm birth at 48 hours and for pneumonia weekly during the first month of life. This will provide a foundation for a larger study aimed at reducing infant morbidity and mortality through the dissemination of a clean gas stove, complemented by a behavioral intervention.